W3C home > Mailing lists > Public > public-media-annotation@w3.org > October 2008

Re: Question on API paper

From: Felix Sasaki <fsasaki@w3.org>
Date: Thu, 16 Oct 2008 21:36:46 +0900
Message-ID: <48F7355E.8030109@w3.org>
To: Raphaël Troncy <Raphael.Troncy@cwi.nl>
CC: public-media-annotation@w3.org

Thank you for the explanation, Raphael.

Raphaël Troncy さんは書きました:
> Dear Felix,
>
>> I have read
>> http://homepages.cs.ncl.ac.uk/phillip.lord/download/publications/cooking03.pdf 
>>
>> and have a general question about the paper: how is it related to the
>> API which we are supposed to produce?
>
> OWL is indeed a language for representing ontologies. The OWL API is 
> therefore the programmatic access to manipulate an ontology, i.e. 
> creating classes, properties and individuals. I think it is relevant 
> because this is an example of building an API on top of a model.
>
>> The paper describes - if I
>> understand correctly - an API to access ontologies, e.g. query, update,
>> serialization etc. However I think we are supposed to create an API for
>> accessing *metadata*, and our API relies on an ontology (thinking of
>> issues 6113, we still have find out how the relation is). I am probably
>> missing something, so many thanks in advance for the clarification.
>
> Well, this is the difference between classes and instances.
> The OWL API allows to access, create, delete, modify classes, 
> properties and individuals of an ontology. The API we should design 
> should allow to access, create, delete, modify metadata, that is, 
> instances of classes and properties defined in the 'Media Annotation 
> Ontology' (individuals in the OWL terminology). The overlap is thus 
> that both API handle individuals.

I know this sounds probably boring, but I still have
http://lists.w3.org/Archives/Public/public-media-annotation/2008Sep/0045.html
and issue 6113 in mind, where Dave pointed out that having a small set 
of tags as a result of our work would be useful. In terms of the API, 
for me that would translate to an API that may be related to the 
ontology, but "must" (as a very strong requirement) be useable without  
any  relation to the ontology. If we want to have widespread adoption in 
the browser community, the ordinary web developer needs to be able to 
execute the operations we are talking about (query, update, ...) without 
any knowledge  about the ontology. Do you think with the API you are 
having in mind that is feasible?

Felix
Received on Thursday, 16 October 2008 12:37:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 16 October 2008 12:37:28 GMT